Rapid manual chest compression (120 compressions/min) CPR has been shown to improve hemodynamics and survival when compared with standard CPR (60 compressions/min) in a canine model of prolonged cardiac arrest. The study showing improved survival with rapid manual CPR empirically included treatment with bicarbonate and initial fluid loading. To determine the role of bicarbonate and fluid loading in the success of rapid manual chest compression CPR, 31 mongrel dogs were studied. After instrumentation with micromanometer-tipped catheters to measure aortic and right atrial pressures, the animals were assigned sequentially to three treatment groups. Group A underwent rapid manual chest compressions at 120 compressions/min, bicarbonate treatment, and initial fluid loading. Group B underwent rapid manual compressions at 120 compressions/min without bicarbonate or fluid loading. Group C underwent standard CPR at 80 compressions/min with bicarbonate and fluid loading. After 30 minutes of ventricular fibrillation, defibrillation was attempted. Seven of 11 dogs in group A survived 24 hours. None of the animals in group B resuscitated or survived. Three of the ten dogs in group C survived 24 hours. Survival with rapid manual CPR without bicarbonate and initial fluid loading was significantly less than when these interventions were used ( P < .01). To examine the separate contribution of bicarbonate and fluid therapy, two additional groups of animals were studied. Fourteen animals (group D) received rapid manual CPR with bicarbonate therapy, and 12 (group E) received rapid manual CPR with fluid loading only. Three of 14 in group D and two of 12 in group E survived 24 hours. This study confirms the benefit of using rapid manual chest compression CPR compared with standard CPR. However, use of bicarbonate and fluid loading is necessary to achieve improved outcome with rapid manual chest compression CPR.